Metraipolis › archive › dc_e85e5071bbc3

Do As I Do, Not As I Say: On-Policy Distillation Inherits the Teacher in Vision-Language Models — arXiv:2610.20989

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.20989
License (as archived)
CC0-1.0
Archived
2026-10-16T10:20:06.624Z
SHA-256 of the archived text
313970f8ddfc088d293038e1341759afc67a81d035ede73d95185df623236b84
Stable URL of this page
https://metraipolis.com/documento/dc_e85e5071bbc3

How to cite

Do As I Do, Not As I Say: On-Policy Distillation Inherits the Teacher in Vision-Language Models — arXiv:2610.20989 — arXiv (en), https://arxiv.org/abs/2610.20989 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_e85e5071bbc3 (Metraipolis archive dc_e85e5071bbc3, sha256 313970f8ddfc088d293038e1341759afc67a81d035ede73d95185df623236b84)
BibTeX
@misc{dce85e5071bbc3,
  title = {Do As I Do, Not As I Say: On-Policy Distillation Inherits the Teacher in Vision-Language Models — arXiv:2610.20989},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.20989},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_e85e5071bbc3; sha256 313970f8ddfc088d293038e1341759afc67a81d035ede73d95185df623236b84},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.20989. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Improving the reasoning capabilities of vision-language models (VLMs) often relies on post-training with supervision from stronger models. Two common strategies are supervised fine-tuning (SFT), which trains on teacher-generated demonstrations, and on-policy distillation (OPD), which provides teacher feedback on trajectories generated by the student itself. While OPD can be more data-efficient by supervising student-generated trajectories, it is unclear how the teacher's supervision at each state affects the student's learned behavior. To study this question, we conduct a controlled comparison of OPD and filtered SFT across data budgets and tasks where the teacher exhibits different levels of reliability, and analyze how these differences shape student accuracy and completion behavior. Across the two tasks, we find that OPD is effective only when the teacher remains reliable on student-generated trajectories. On ChartQA, where the teacher completes 99% of its solutions, OPD trained on only 10% of the data comes within 1.4 accuracy points of full-data SFT. In contrast, on Geometry3K, where the teacher fails to complete 22% of its solutions, OPD underperforms SFT by 7.9 points, with most of the gap explained by incomplete responses. More strikingly, starting OPD from an SFT student reduces its completion rate from 76% to 62%, effectively erasing the completion behavior learned through SFT. Filtering teacher demonstrations for completeness alone recovers much of SFT's advantage. These results show that on-policy supervision can transfer not only teacher knowledge but also persistent failure modes on student-visited states. Code is available at this https URL .

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno).
Categorie arXiv: Machine Learning (cs.LG) ; Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2610.20989 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-16T08:48:02.668Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).